DPC Latency and CH Products

[bustr]

I've just upgraded from a X52 Pro to a CH Pro Throttle USB and CH Fhighter Stick USB.

The CH products are on an external powered USB HUB. At idle my system now runs at 3-4% CPU Usage on both CPU. I have a Core Duo E8500. Using RATTV3 from MS, Process Explorer from Sysinternals and DPC Latency Checker I discoverd the source of my DPC latency was the CH products. If I unplugged each at a time the latency dropped from 500us to 250us to a base level of 7-2us which registers as 0% CPU usage in task manager.

With my Saitek X52 Pro there was never a DPC latency issue at CPU idle. Is this anything I should be concerned with? Is this happening because I have the CH Control Manager installed? I don't think it is the external Belkin HUB. If I pull all devices from it, the CPU Usage drops to 0%.

If anyone would like to check their own DPC latency, DPC Latency Checker can be downloaded from here:  http://www.thesycon.de/eng/latency_check.shtml

RATTV3 can be downloaded from here and will show you the driver(s) causing high DPC Latency:  http://www.microsoft.com/whdc/DevTools/tools/RATT.mspx

Process Explorer from Sysinternals will show you if you are using DPC's. Task Manager under performance shows this as CPU Usage:  http://technet.microsoft.com/en-us/sysinternals/bb896653.aspx

If you are having CPU usage issues at idle, these appletts may help you find the source.

[bustr]

I just analysed my USB bus traffic with Bus Hound by Perisoft. The 4% CPU usage at idle or 500us DPC Latency is due to a small amount of potentiometer spiking being picked up by the system as data transfers from the CH Products. You can down load Bus Hound and run the capture option. If your controllers are spiking while at idle you will capture transfer data along with the source driver being listed. Bus Hound is not freeware but, it will install and run in a limited mode. Thats just enough to capture USB traffic to verify if your 2-4% CPU activity is from your game controllers.

I've seen repeated threads from players stating they have reduced their running services and processes to the minimum Skuzzy reccommends for playing the game but, still reporting 2-4% CPU usage at idle. Many have responded to reboot in safe mode and perform extensive virus checking. Is it possible for a number of players that 2-4% CPU Usage is coming from DPC Latency realated to their USB game controler's potentiometers randomly spiking at idle?

[maddafinga]

Well, I have CH gear and that same 2-4% usage you describe exactly.  I never really worried about it or tried to track it down though.  Since you have the control manager installed, do you think it would make any difference to put a bit of a software deadband on?  Just an idea, no clue as to whether or not it would work.

[Auger]

Is it spiking or pot noise getting picked up by the CH drivers?  I noticed the CPU utilization caused by the CH sticks, but I don't think it has ever been an issue.  And it is very unlikely that I'm going to notice a 1/2 millisecond of latency in anything I do on my Windows system.

But that is excellent research.  I always wondered why I had up to 10% CPU utilization on XP with a whopping 19 processes running and the CH sticks plugged in. 

[bustr]

The Pro Throttle has spiking on the throttle and the mini-joystick. The Fighterstick is steady on X,Y & Z axis. If I pull the throttle it cuts my DPC Latency by half. 500us down to 250us. The CH products have programable buttons. You can see active power LED "on" and green when they are sitting idle on the USB bus. Probably only CH can answer the following question:

Is it possible the CH products I have installed create the bus activity due to the nature of their onboard HID components?

Even with this minor CPU utilization, my CH Joystick compaired to 10 years of Saiteks is like learning to drive fast on I-95 between Washington D.C. and Richamond VA. Then being put in a Porsche 997 on the Authoban.

Since there are so many different joystick types being used in this game, maybe gentelmen and ladies, you could download DPC Latency Checker and leave a note here about your findings. It may have the positive result of helping players determine if it's USB activity or a virus causing their low end CPU Utilisation at idle.

[Chalenge]

I think this problem goes much deeper than just the CH gear. You didnt say which OS you are using...

Nvidia chipsets and soundblaster cards have a known 'DPC latency interference pattern' that is much worse than the 'normal' onboard sound interference pattern for instance. On Vista it is impossible to get rid of with any Creative device present and on XP it can show up in any of many assorted ways. I am nearly bald from pulling my hair out during experiments to hunt down problems like this. I know Nvidia is at least aware of it but I am not sure what they are doing about it and I have not yet tested W7.

[Shamus]

I am running CH hotas, xp, Intel chipset, usb headphones and getting a DSP of 235 with the CH software running or not showing cpu usage of 2-4% at idle.

shamus 

[RTHolmes]

E8400, X52 pro: 0-3% cpu, 2-70us DPC latency

[Chalenge]

Two test systems

Vista x64, CH Fighterstick Pro, Pro Throttle, Throttle Quadrant, Pro Pedals (all USB), stock clocks E8400 cpu (0-6%), HT Omega Striker 7.1, EVGA 780i, EVGA 8800 GTs (2 in SLI), 8 GB DDR2

Sticks plugged in: 43-2403
Sticks unplugged: 19-63

Vista x64, same joysticks, stock clocks i7-920 Bloomfield (0-7%), HT Omega Striker 7.1, EVGA E760 Classified (Intel x58), EVGA GTX 295 (2 in SLI), 24 GB DDR3

Sticks plugged in: 43-235
Sticks unplugged: 19-47

[llama]

OK great, you're measuring DPC Latency. Congratulations.

I don't hear you mentioning anything about your system being slow, unstable, or with a bad framerate. If your PC is otherwise working fine and the game works trouble-free, then stop worrying about DPC Latency. You can go crazy chasing down computer gophers that really aren't a problem...

-Llama

[Chalenge]

Thanks llama but when it IS a problem (try throwing a SB card in with an nvidia chipset) the latency never drops and stutters dont either.

[bustr]

I disabled my SB F-Xi sound card in favor of a Logitec USB head set for sound. When I pull my CH throttle and joystick CPU Usage = 0% while the Logitec headset\sound chip is still plugged into the Belkin USB HUB.

I performed this debug as a curiosity for the most part, and to give players something more to check than the usual: send a dixdiag, reboot in safemode to run antivirus\spyware detector, or even reinstall your OS.

My last IT position was with Bank of America PES (Problem Escalation Services) and DSS (Distributed Server Support). Anyone here who wants to check into this needs to tell me who D.R. Shaw, Darrel Zink and Mark Carter are in those organisations. I supported 3000 servers from single instance DOS apps to ESX Virtualisation with SAN backend storage using WWN for its routing and security model through McData directors on Hitachi and EMC storage arrays. Many of our issues required debugging in static or live mode while working with our Microsft contract. After all, $100k-$1M loss a day for an outage is reason enough to learn how to do your job. So yes, I know how to debug the DPC Latency and USB bus as a more likely reason some players cannot get their CPU Usage below 2-12% at idle.

I thought I was helping people chalenge and llama. Guess you guys are the big sticks in this rice bowl. I've wondered why over the years players are not directed to very simple to use free tools that will show them much of whats running in their sysjtems stacks and won't damage those systems in the process. Sysinternals has quite a large number of them and MS support usualy directs their contracts to download them to assist in basic debug and problem solveing.

Just for you Chalenge, and by the way, my system is very healthy. It's at 18 processes before I start the game. Vista and Win7 users report higher DPC Latency numbers at idle from their USB devices.

 Machine name: BUSTR
   Operating System: Windows XP Professional (5.1, Build 2600) Service Pack 3 (2600.xpsp_sp3_gdr.080814-1236)
           Language: English (Regional Setting: English)
System Manufacturer: System manufacturer
       System Model: System Product Name
               BIOS: BIOS Date: 12/03/08 11:51:39 Ver: 08.00.14
          Processor: Intel(R) Core(TM)2 Duo CPU     E8500  @ 3.16GHz (2 CPUs)
             Memory: 3326MB RAM
          Page File: 159MB used, 5051MB available
        Windows Dir: C:\WINDOWS
    DirectX Version: DirectX 9.0c (4.09.0000.0904)
DX Setup Parameters: Not found
     DxDiag Version: 5.03.2600.5512 32bit Unicode

------------
DxDiag Notes
------------
  DirectX Files Tab: No problems found.
      Display Tab 1: No problems found.
        Sound Tab 1: No problems found.
          Music Tab: No problems found.
          Input Tab: No problems found.
        Network Tab: No problems found.

--------------------
DirectX Debug Levels
--------------------
Direct3D:    0/4 (n/a)
DirectDraw:  0/4 (retail)
DirectInput: 0/5 (n/a)
DirectMusic: 0/5 (n/a)
DirectPlay:  0/9 (retail)
DirectSound: 0/5 (retail)
DirectShow:  0/6 (retail)

---------------
Display Devices
---------------
        Card name: NVIDIA GeForce 9800 GT 
     Manufacturer: NVIDIA
        Chip type: GeForce 9800 GT
         DAC type: Integrated RAMDAC
       Device Key: Enum\PCI\VEN_10DE&DEV_0614&SUBSYS_C9753842&REV_A2
   Display Memory: 512.0 MB
     Current Mode: 1680 x 1050 (32 bit) (60Hz)
          Monitor: LG L227W (Digital)
  Monitor Max Res: 1680,1050
      Driver Name: nv4_disp.dll
   Driver Version: 6.14.0011.8618 (English)
      DDI Version: 9 (or higher)
Driver Attributes: Final Retail
 Driver Date/Size: 6/10/2009 06:03:00, 5908608 bytes
      WHQL Logo'd: Yes
  WHQL Date Stamp: n/a
              VDD: n/a
         Mini VDD: nv4_mini.sys
    Mini VDD Date: 6/10/2009 06:03:00, 8087712 bytes
Device Identifier: {D7B71E3E-4554-11CF-587A-7EE903C2CB35}
        Vendor ID: 0x10DE
        Device ID: 0x0614
        SubSys ID: 0xC9753842
      Revision ID: 0x00A2
      Revision ID: 0x00A2
      Video Accel:
 Deinterlace Caps: n/a
         Registry: OK
     DDraw Status: Enabled
       D3D Status: Enabled
       AGP Status: Disabled
DDraw Test Result: Not run
 D3D7 Test Result: Not run
 D3D8 Test Result: Not run
 D3D9 Test Result: Not run

-------------
Sound Devices
-------------
            Description: Logitech USB Headset
 Default Sound Playback: Yes
 Default Voice Playback: Yes
            Hardware ID: USB\Vid_046d&Pid_0a0b&Rev_1013&MI_00
        Manufacturer ID: 65535
             Product ID: 65535
                   Type: WDM
            Driver Name: usbaudio.sys
         Driver Version: 5.01.2600.5512 (English)
      Driver Attributes: Final Retail
            WHQL Logo'd: Yes
          Date and Size: 4/14/2008 00:15:14, 60032 bytes
            Other Files:
        Driver Provider: Microsoft
         HW Accel Level: Standard
              Cap Flags: 0xF5F
    Min/Max Sample Rate: 100, 200000
Static/Strm HW Mix Bufs: 0, 0
 Static/Strm HW 3D Bufs: 0, 0
              HW Memory: 0
       Voice Management: No
 EAX(tm) 2.0 Listen/Src: No, No
   I3DL2(tm) Listen/Src: No, No
Sensaura(tm) ZoomFX(tm): No
               Registry: OK
      Sound Test Result: Not run

---------------------
Sound Capture Devices
---------------------
            Description: Logitech USB Headset
  Default Sound Capture: Yes
  Default Voice Capture: Yes
            Driver Name: usbaudio.sys
         Driver Version: 5.01.2600.5512 (English)
      Driver Attributes: Final Retail
          Date and Size: 4/14/2008 00:15:14, 60032 bytes
              Cap Flags: 0x41
           Format Flags: 0x444

[Chalenge]

I wasnt trying to put down your information bustr but I was pointing out what I have learned since the time I had a soundblaster card and came across the latency app. I have no credentials to share with you whatsoever so please keep contributing.

I also discovered with this app that the nvidia chipset drivers and probably the video drivers too are not playing nicely and do cause latency issues themselves. Yes the USB devices (CH) cause an increase in latency but I dont think they are the underlying problem with spiking (correct me if Im wrong).

[Chalenge]

I went through my processes and aggressively shut down everything that is not required or that I could recognize as not required (got Vista down to 46 processes). In doing that I shutdown Protexis Licensing (required for Paint Shop Pro x2) and FNPlicensingservice (Macrovision) and several True Image processes and now even with the CH sticks plugged in my latency is reported at less than 70 and usually hovers around 46-50.

[Wing19]

I run CH gear and I'm at 0 to 2% with 21 process running but I only use this system for gaming